CN113057617A - Non-invasive monitoring system for cardiac output - Google Patents
Non-invasive monitoring system for cardiac output Download PDFInfo
- Publication number
- CN113057617A CN113057617A CN202110482361.4A CN202110482361A CN113057617A CN 113057617 A CN113057617 A CN 113057617A CN 202110482361 A CN202110482361 A CN 202110482361A CN 113057617 A CN113057617 A CN 113057617A
- Authority
- CN
- China
- Prior art keywords
- module
- waveform
- blood pressure
- cardiac output
- arterial blood
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/026—Measuring blood flow
- A61B5/029—Measuring or recording blood output from the heart, e.g. minute volume
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/021—Measuring pressure in heart or blood vessels
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Signal Processing (AREA)
- Physiology (AREA)
- Animal Behavior & Ethology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Cardiology (AREA)
- Hematology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Power Engineering (AREA)
- Vascular Medicine (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The invention discloses a non-invasive monitoring system of cardiac output, which comprises a data acquisition control module, a non-invasive peripheral arterial blood pressure measuring module, a waveform screening module, a data preprocessing module, a cardiac output calculating module and a data display module; the data acquisition control module is used for controlling the data acquisition and data transmission processes; the waveform screening module is used for classifying and screening peripheral arterial blood pressure waveforms to screen out A-type waveforms; the data preprocessing module is used for preprocessing the A-type waveform and sending the A-type waveform to the cardiac output calculating module; the cardiac output calculating module comprises a neural network trained and completed through a sample set, the trained neural network has end-to-end identification capability, namely peripheral arterial blood pressure waveforms are input and corresponding cardiac output is output, and the cardiac output is sent to the data display module through the data acquisition control module. The invention can perform end-to-end identification and can perform noninvasive continuous monitoring of cardiac output.
Description
Technical Field
The invention belongs to the technical field of cardiac output measurement, and particularly relates to the technical field of noninvasive cardiac output measurement.
Background
The Cardiac Output (CO) is the ejection volume of the left ventricle per minute, is the most important parameter for characterizing the health status of the cardiovascular system, and is an important diagnostic basis for cardiac function and cardiovascular diseases. In addition, many other cardiovascular system parameters can be calculated in an auxiliary manner based on the cardiac output, so that accurate measurement of the cardiac output is very critical in the aspects of cardiovascular disease detection, treatment and the like, and has important clinical significance.
Currently, devices and methods for measuring cardiac output fall into three major categories, invasive, minimally invasive and non-invasive. Compared with the human body, invasive and minimally invasive measurement technologies have injury, cannot continuously act on the human body for long-time continuous monitoring, and are generally suitable for single measurement or short-time measurement. The non-invasive measurement mainly comprises an ultrasonic method, a thoracic impedance method and a pulse wave waveform analysis method. The ultrasonic method needs to emit ultrasonic waves to a human body, and the thoracic impedance method needs to apply oscillating current to the human body, so that the method is not suitable for continuously monitoring the human body. The pulse waveform contains abundant physiological information such as heart rate, average pressure, systolic pressure, arteriosclerosis, peripheral resistance, reflected wave intensity, and the like. Some scholars propose methods for measuring CO by using pulse wave characteristics, but most of the methods are limited to manually extracting relevant characteristics from pulse waves, then calculating stroke volume according to the characteristics, and then calculating cardiac output according to the stroke volume, so that the measurement efficiency is low, manual intervention is needed, continuous monitoring for 24 hours cannot be achieved, and the measurement accuracy is still required to be improved due to the influence of subjective factors.
Disclosure of Invention
Aiming at the defects of the technology, the invention provides a non-invasive monitoring system of cardiac output, which solves the technical problem of how to realize the non-invasive continuous monitoring of the cardiac output.
In order to solve the technical problems, the technical scheme of the invention is as follows: a non-invasive monitoring system for cardiac output comprises a data acquisition control module, a non-invasive peripheral arterial blood pressure measurement module, a waveform screening module, a data preprocessing module, a cardiac output calculation module and a data display module;
the data acquisition control module is used for controlling the data acquisition and data transmission processes;
the peripheral arterial blood pressure measuring module is used for collecting peripheral arterial blood pressure waveforms under the control of the data collection control module and sending the peripheral arterial blood pressure waveforms to the waveform screening module through the data collection control module;
the waveform screening module is used for classifying and screening the peripheral arterial blood pressure waveforms, only the current peripheral arterial blood pressure waveforms belong to A-type waveforms, and the current peripheral arterial blood pressure waveforms are sent to the data and preprocessing module through the data acquisition control module;
the A-type waveform has complete cardiac cycle and waveform typical characteristics, and partial wave bands have power frequency interference, baseline drift and transient waveform delay;
the data preprocessing module is used for preprocessing the A-type waveform, filtering out basic noise by adopting smooth filtering, Butterworth band-pass filtering and low-pass filtering, and performing waveform decomposition and reconstruction by adopting wavelet transformation and EMD (empirical mode decomposition) so as to filter out baseline drift and the like; the peripheral arterial blood pressure waveform processed and completed by the data preprocessing module is sent to the cardiac output calculating module through a data acquisition control module;
the cardiac output calculation module comprises a neural network trained by a sample set, wherein the samples in the sample set are constructed as follows: peripheral arterial blood pressure waveforms collected in a non-invasive mode are used as input, the cardiac output of the same patient collected in an invasive mode is used as a label, the trained neural network has end-to-end recognition capability, namely the peripheral arterial blood pressure waveforms are input and corresponding cardiac output is output, and the cardiac output is sent to the data display module through the data collection control module.
The data acquisition control module controls the heart rate module to acquire the heart rate according to the beat amount calculation request of the beat amount calculation module and sends the heart rate and the cardiac output quantity to the beat amount calculation module, and the beat amount calculation module is used for calculating the beat amount according to the cardiac output quantity and the heart rate.
Further, the peripheral arterial blood pressure waveform finished by the data preprocessing module is respectively sent to the data display module and the cardiac output calculating module through the data acquisition control module.
Further, the neural network comprises a one-dimensional convolution neural network and an LSTM bidirectional long-time memory network; the one-dimensional convolutional neural network is used for extracting a feature vector of the preprocessed peripheral arterial blood pressure waveform, performing dimension conversion on the feature vector through a convolutional layer and a pooling layer to adapt to the input of the LSTM bidirectional long-time memory network, and keeping the periodicity of the peripheral arterial blood pressure waveform; the LSTM bidirectional long-time and short-time memory network is used for splicing the characteristic vectors along the time dimension and performing regression prediction on the cardiac output.
Further, if the current peripheral arterial blood pressure waveform belongs to a B-type waveform or a C-type waveform, sending alarm information to the data acquisition control module, sending the alarm information to the data display module by the data acquisition control module, and controlling the peripheral arterial blood pressure measurement module to repeatedly execute alternate pause and start until the A-type waveform is detected;
the B-type waveform has a complete cardiac cycle, but the noise interference is too large, the waveform is not in accordance with the form of a normal waveform, and the waveform jitter amplitude is not in the range of human body vital signs; the class C waveform does not have a complete cardiac cycle; the forming reasons of the B-type waveform and the C-type waveform comprise electrode falling and peripheral arterial blood pressure waveform collection under the unsmooth state of a measured object.
Further, if the current peripheral arterial blood pressure waveform does not belong to any one of the A-type waveform, the B-type waveform and the C-type waveform, alarm information is sent to the data acquisition control module, and the data acquisition control module sends the alarm information to the data display module and controls the peripheral arterial blood pressure measurement module to be closed to stop working.
Compared with the prior art, the invention has the advantages that:
1. the invention overcomes the technical bottleneck problems of great difficulty in artificial feature extraction, incomplete characterization and the like. The method realizes the end-to-end prediction from the peripheral arterial blood pressure waveform to the cardiac output, eliminates abnormal waveforms by classifying and screening the waveforms, avoids inputting the abnormal waveforms into a neural network during training or running, improves the performance of the neural network, and improves the accuracy of end-to-end identification by the support of preprocessing comprehensive technology on the waveforms.
2. The neural network is an end-to-end black box model, and the whole identification process does not need manual intervention, so that the method can be used for continuously monitoring the cardiac output for 24 hours.
3. The method combining the convolutional neural network and the long-term memory network is utilized to realize automatic extraction of high-dimensional characteristics of input data, reveal close relation between cardiac output and arterial blood pressure waveform, improve the generalization capability of the model, improve the calculation precision of the cardiac output and effectively reduce measurement errors.
Drawings
FIG. 1 is a flow chart of a system for non-invasive monitoring of cardiac output for cardiac output prediction;
FIG. 2 is a schematic diagram of feature vector extraction for a one-dimensional convolutional neural network;
FIG. 3 is a diagram of the elements of an LSTM network;
FIG. 4 is a graph of the predicted effect of a non-invasive cardiac output monitoring system on a training set;
fig. 5 is a graph of the predicted effect of a non-invasive cardiac output monitoring system on a test set.
Detailed Description
A non-invasive monitoring system for cardiac output comprises a data acquisition control module, a non-invasive peripheral arterial blood pressure measurement module, a waveform screening module, a data preprocessing module, a cardiac output calculation module and a data display module; the data acquisition control module is used for controlling the data acquisition and data transmission processes.
The non-invasive peripheral arterial blood pressure measuring module is used for collecting peripheral arterial blood pressure waveforms under the control of the data collection control module and sending the peripheral arterial blood pressure waveforms to the waveform screening module through the data collection control module.
The non-invasive peripheral arterial blood pressure measuring module comprises a piezoelectric sensor or a photoelectric sensor, and pulse signals obtained by the piezoelectric sensor or the photoelectric sensor are transmitted to a blood pressure signal processing circuit through a lead to be filtered and amplified to form a peripheral arterial blood pressure waveform. Peripheral arterial blood pressure waveforms at the brachial, radial or carotid arteries can be collected.
The waveform screening module is used for classifying and screening the peripheral arterial blood pressure waveforms, only the current peripheral arterial blood pressure waveforms belong to A-type waveforms, and the current peripheral arterial blood pressure waveforms are sent to the data and preprocessing module through the data acquisition control module; the A-type waveform has complete cardiac cycle and waveform typical characteristics, and partial wave bands have power frequency interference, baseline drift and transient waveform delay;
if the current peripheral arterial blood pressure waveform belongs to a B-type waveform or a C-type waveform, sending alarm information to a data acquisition control module, sending the alarm information to a data display module by the data acquisition control module, and controlling a peripheral arterial blood pressure measurement module to repeatedly execute alternate pause and start until a type A waveform is detected;
the B-type waveform has a complete cardiac cycle, but the noise interference is too large, the waveform is not in accordance with the form of a normal waveform, the waveform jitter amplitude is seriously distorted, and the waveform is not in the range of human vital sign parameters; the C-type waveform has no complete cardiac cycle or unobvious periodic characteristics, and the forming reasons of the B-type waveform and the C-type waveform comprise electrode falling and peripheral arterial blood pressure waveform acquisition under the state that a measured object is not static.
If the current peripheral arterial blood pressure waveform does not belong to any one of the A-type waveform, the B-type waveform and the C-type waveform, alarm information is sent to the data acquisition control module, and the data acquisition control module sends the alarm information to the data display module and controls the peripheral arterial blood pressure measurement module to be closed to stop working.
Waveform classification and identification: waveform classification and identification: after the pulse wave waveform is divided into single cycles, Signal Labeler APP of matlab2020a is used for labeling the pulse wave waveform, namely A-type waveforms, B-type waveforms and C-type waveforms respectively, the x axis and the y axis of the pulse wave waveform Signal and the differential pulse wave Signal are removed, then the two signals are combined and converted into pictures, the pictures are input into a two-dimensional depth convolution neural network, and finally 3 types of waveforms are output.
The data preprocessing module is used for preprocessing the A-type waveform, filtering basic noise by adopting smooth filtering, Butterworth band-pass filtering and low-pass filtering, and performing waveform decomposition and reconstruction by adopting wavelet transformation and EMD (empirical mode decomposition) so as to filter baseline drift and the like; the peripheral arterial blood pressure waveform preprocessed by the data preprocessing module is sent to the cardiac output calculating module through a data acquisition control module or respectively sent to the data display module and the cardiac output calculating module, and the data display module displays the preprocessed peripheral arterial waveform.
The cardiac output calculation module includes a neural network trained through a sample set, the samples in the sample set being constructed as follows: peripheral arterial blood pressure waveforms collected in a non-invasive mode are used as input, the cardiac output of the same patient collected in an invasive mode is used as a label, the trained neural network has end-to-end recognition capability, namely the peripheral arterial blood pressure waveforms are input and corresponding cardiac output is output, and the cardiac output is sent to the data display module through the data collection control module.
During training, a sample is input from the waveform screening module, the sample set is divided into a training set and a testing set, and if the prediction effect of the testing set meets the requirement, the neural network training is finished.
The neural network comprises a one-dimensional convolution neural network and an LSTM bidirectional long-time memory network; the one-dimensional convolutional neural network is used for extracting a feature vector of the preprocessed peripheral arterial blood pressure waveform, performing dimension conversion on the feature vector through a convolutional layer and a pooling layer to adapt to the input of the LSTM bidirectional long-time memory network, and keeping the periodicity of the peripheral arterial blood pressure waveform; the LSTM bidirectional long-time and short-time memory network is used for splicing the characteristic vectors along the time dimension and performing regression prediction on the cardiac output.
Referring to fig. 2, the one-dimensional convolutional neural network extracts morphological features of a blood pressure signal, retains the periodicity of an original signal, overcomes the defects of manual feature extraction and high clinical use error of the traditional cardiac output prediction through a convolutional layer and a pooling layer, and improves measurement accuracy by using an arterial blood pressure waveform with periodicity to perform end-to-end frontal cardiac output measurement. The specific steps of extracting the feature vector by the one-dimensional convolutional neural network are as follows:
firstly, extracting features through a one-dimensional convolution neural network, convolving an input one-dimensional blood pressure signal with a one-dimensional convolution kernel by a CNN convolution layer, and then outputting the features through Relu function activation, maximum pooling and the like, wherein the convolution kernel of each filter uses the same convolution kernel to extract signal features:
yk+1,m(n)=wk,m*xk(n)+bk,m (1)
in the formula (1), wk,mAnd bk,mRespectively representing the weight and the offset of the mth filter core in the kth layer; x is the number ofk(n) denotes an nth signal in a kth layer; y is(k+1,m)The output of the m-th filtered kernel convolution representing the nth signal of the (k + 1) -th layer. After convolution, the feature expression capability of the model is enhanced by activating a function, and the convergence of the model is accelerated by using a Relu activation function, such as a formula
ak+1,m(n)=max{0,yk+1,m(n)} (2)
A in formula (2)k+1,m(n) is yk+1,m(n) the activated output value. Finally, we adopt the maximum pooling layer to reduce the network space and feature dimension after CNN convolution, and the maximum pooling operation is as follows:
in formula (3): o isk,m(t) represents the output of the t-th neuron in the K-th layer, OK+1,m(n) represents the output value after pooling.
After CNN, obtaining different convolution kernels and multi-channel characteristic time sequence signals after pooling, and sending the signals into an LSTM processing unit as the input of an LSTM network.
Structure diagram of elements of LSTM network referring to fig. 3, the LSTM network predicts as follows:
1) the cell state is multiplied by the output of this activation function (sigmoid) by the input of the current time and the output of the previous time hidden layer via the sigmoid function. If the output is 0, the part of information needs to be forgotten, and the current information continues to be transmitted in the unit state.
It=S(Whfht-1+WifXt) (4)
2) The old cell state is updated. The previous forgetting threshold layer determines which information is forgotten or added, and the threshold layer is used for inputting and executing all information.
Nt=δ(Whiht-1+WliXt) (5)
Ut=tanh(Whmht-1+WiXt) (6)
3) The unit state corresponds to a conveyor belt, on which the contents increase or decrease as he passes through each repeat module based on the current input.
Ct=MtCt-1+NiUt (7)
The full connection layer comprises an LSTM network, vectors output by an LSTM unit are connected according to weights, the function of dimension conversion is realized, and finally, a cardiac output value is output.
The system of the invention is used for testing in clinical experiments of 202 patients, and the end-to-end cardiac output measurement is innovatively used, so that the effect is ideal. In clinical experiments, 202 patients have cardiac output values measured by invasive catheter interventional devices, arterial blood pressure waveforms of the non-invasively acquired patients correspond to the existing cardiac output values (labels) one by one, the arterial blood pressure waveforms are input into the system, predicted cardiac output values are obtained, and the effect comparison graph of partial cardiac output values is shown as follows: in the figure, the horizontal axis represents the real value of the input system, the vertical axis represents the predicted value of the output, and with reference to fig. 4 and 5, the accuracy on the training set is 0.90759, and the accuracy on the test set is 0.71284. It can be seen that the overall effect of the model is better.
Claims (7)
1. A system for non-invasive monitoring of cardiac output, comprising: the device comprises a data acquisition control module, a non-invasive peripheral arterial blood pressure measurement module, a waveform screening module, a data preprocessing module, a cardiac output calculation module and a data display module; the data acquisition control module is used for controlling the data acquisition and data transmission processes;
the non-invasive peripheral arterial blood pressure measuring module is used for collecting peripheral arterial blood pressure waveforms under the control of the data collection control module and sending the peripheral arterial blood pressure waveforms to the waveform screening module through the data collection control module;
the waveform screening module is used for classifying and screening the peripheral arterial blood pressure waveforms, only the current peripheral arterial blood pressure waveforms belong to A-type waveforms, and the current peripheral arterial blood pressure waveforms are sent to the data and preprocessing module through the data acquisition control module;
the A-type waveform has complete cardiac cycle and waveform typical characteristics, and partial wave bands have power frequency interference, baseline drift and transient waveform delay;
the data preprocessing module is used for preprocessing the A-type waveform, filtering out basic noise by adopting smooth filtering, Butterworth band-pass filtering and low-pass filtering, and performing waveform decomposition and reconstruction by adopting wavelet transformation and EMD (empirical mode decomposition) so as to filter out baseline drift and the like; the peripheral arterial blood pressure waveform preprocessed by the data preprocessing module is sent to the cardiac output calculating module through a data acquisition control module;
the cardiac output calculation module comprises a neural network trained by a sample set, wherein the samples in the sample set are constructed as follows: peripheral arterial blood pressure waveforms collected in a non-invasive mode are used as input, the cardiac output of the same patient collected in an invasive mode is used as a label, the trained neural network has end-to-end recognition capability, namely the peripheral arterial blood pressure waveforms are input and corresponding cardiac output is output, and the cardiac output is sent to the data display module through the data collection control module.
2. A system for non-invasive monitoring of cardiac output according to claim 1, characterized in that: the data acquisition control module controls the heart rate module to acquire the heart rate according to the beat amount calculation request of the beat amount calculation module and sends the heart rate and the cardiac output quantity to the beat amount calculation module, and the beat amount calculation module is used for calculating the beat amount according to the cardiac output quantity and the heart rate.
3. A system for non-invasive monitoring of cardiac output according to claim 1, characterized in that: the peripheral arterial blood pressure waveform preprocessed by the data preprocessing module is respectively sent to the data display module and the cardiac output calculating module through the data acquisition control module.
4. A system for non-invasive monitoring of cardiac output according to claim 1, characterized in that: the neural network comprises a one-dimensional convolution neural network and an LSTM bidirectional long-time memory network; the one-dimensional convolutional neural network is used for extracting a feature vector of the preprocessed peripheral arterial blood pressure waveform, performing dimension conversion on the feature vector through a convolutional layer and a pooling layer to adapt to the input of the LSTM bidirectional long-time memory network, and keeping the periodicity of the peripheral arterial blood pressure waveform; the LSTM bidirectional long-time and short-time memory network is used for splicing the characteristic vectors along the time dimension and performing regression prediction on the cardiac output.
5. A system for non-invasive monitoring of cardiac output according to claim 1, characterized in that: the non-invasive peripheral arterial blood pressure measuring module comprises a piezoelectric sensor or a photoelectric sensor, and pulse signals obtained by the piezoelectric sensor or the photoelectric sensor are transmitted to a blood pressure signal processing circuit through a lead to be filtered and amplified to form a peripheral arterial blood pressure waveform.
6. A system for non-invasive monitoring of cardiac output according to claim 1, characterized in that: if the current peripheral arterial blood pressure waveform belongs to a B-type waveform or a C-type waveform, sending alarm information to a data acquisition control module, sending the alarm information to a data display module by the data acquisition control module, and controlling a peripheral arterial blood pressure measurement module to repeatedly execute alternate pause and start until a type A waveform is detected;
the B-type waveform has a complete cardiac cycle, but the noise interference is too large, the waveform does not conform to the form of a normal waveform, and the waveform jitter amplitude is not in the range of human vital sign parameters; the class C waveform has no complete cardiac cycle, or the periodic characteristics are not apparent; the forming reasons of the B-type waveform and the C-type waveform comprise electrode falling and peripheral arterial blood pressure waveform collection under the unsmooth state of a measured object.
7. System for non-invasive monitoring of cardiac output according to claim 6, characterized in that: if the current peripheral arterial blood pressure waveform does not belong to any one of the A-type waveform, the B-type waveform and the C-type waveform, alarm information is sent to the data acquisition control module, and the data acquisition control module sends the alarm information to the data display module and controls the peripheral arterial blood pressure measurement module to be closed to stop working.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110482361.4A CN113057617B (en) | 2021-04-30 | 2021-04-30 | Non-invasive monitoring system for cardiac output |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110482361.4A CN113057617B (en) | 2021-04-30 | 2021-04-30 | Non-invasive monitoring system for cardiac output |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113057617A true CN113057617A (en) | 2021-07-02 |
CN113057617B CN113057617B (en) | 2022-08-26 |
Family
ID=76568193
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110482361.4A Active CN113057617B (en) | 2021-04-30 | 2021-04-30 | Non-invasive monitoring system for cardiac output |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113057617B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113499048A (en) * | 2021-07-22 | 2021-10-15 | 重庆理工大学 | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM |
Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6358213B1 (en) * | 1999-08-18 | 2002-03-19 | Critikon Company, Llc | Calculation of quality and its use in determination of indirect noninvasive blood pressure |
CN102113878A (en) * | 2009-12-30 | 2011-07-06 | 迈瑞Ds美国有限责任公司 | Monitor and method applied to same |
JP2016131604A (en) * | 2015-01-16 | 2016-07-25 | セイコーエプソン株式会社 | Biological information measurement system, biological information measurement device, and biological information measurement method |
US20170119261A1 (en) * | 2009-04-22 | 2017-05-04 | Streamline Automation Llc | System and Method for Noninvasively Measuring Ventricular Stroke Volume and Cardiac Output |
WO2017129495A1 (en) * | 2016-01-28 | 2017-08-03 | Koninklijke Philips N.V. | Pulse rate measurement module and method |
CN108932452A (en) * | 2017-05-22 | 2018-12-04 | 中国科学院半导体研究所 | Arrhythmia cordis beat classification method based on multiple dimensioned convolutional neural networks |
CN109635291A (en) * | 2018-12-04 | 2019-04-16 | 重庆理工大学 | A kind of recommended method of fusion score information and item contents based on coorinated training |
US10349887B1 (en) * | 2015-06-14 | 2019-07-16 | Facense Ltd. | Blood pressure measuring smartglasses |
US20190313915A1 (en) * | 2015-06-14 | 2019-10-17 | Facense Ltd. | Posture-adjusted calculation of physiological signals |
US20200138306A1 (en) * | 2018-11-02 | 2020-05-07 | Samsung Electronics Co., Ltd. | Feature selection for cardiac arrhythmia classification and screening |
CN111110208A (en) * | 2019-12-13 | 2020-05-08 | 南京理工大学 | LSTM-based oxygen reduction state prediction method for chronic obstructive pulmonary disease |
US20200222607A1 (en) * | 2019-01-16 | 2020-07-16 | Abiomed, Inc. | Left Ventricular Volume and Cardiac Output Estimation Using Machine Learning Model |
CN111493855A (en) * | 2020-04-21 | 2020-08-07 | 重庆理工大学 | System and method for non-invasive measurement of individualized cardiac output |
CN111528814A (en) * | 2020-04-29 | 2020-08-14 | 浙江工业大学 | Method for monitoring blood pressure through machine learning based on LSTM neural network |
WO2020227514A1 (en) * | 2019-05-08 | 2020-11-12 | Lifelens Technologies, Llc | Monitoring and processing physiological signals to detect and predict dysfunction of an anatomical feature of an individual |
CN112071419A (en) * | 2019-06-11 | 2020-12-11 | 西门子医疗有限公司 | Hemodynamic analysis of blood vessels using recurrent neural networks |
CN112274127A (en) * | 2020-10-28 | 2021-01-29 | 河北工业大学 | Noninvasive continuous blood pressure detection method and device based on one-way pulse wave |
-
2021
- 2021-04-30 CN CN202110482361.4A patent/CN113057617B/en active Active
Patent Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6358213B1 (en) * | 1999-08-18 | 2002-03-19 | Critikon Company, Llc | Calculation of quality and its use in determination of indirect noninvasive blood pressure |
US20170119261A1 (en) * | 2009-04-22 | 2017-05-04 | Streamline Automation Llc | System and Method for Noninvasively Measuring Ventricular Stroke Volume and Cardiac Output |
CN102113878A (en) * | 2009-12-30 | 2011-07-06 | 迈瑞Ds美国有限责任公司 | Monitor and method applied to same |
JP2016131604A (en) * | 2015-01-16 | 2016-07-25 | セイコーエプソン株式会社 | Biological information measurement system, biological information measurement device, and biological information measurement method |
US10349887B1 (en) * | 2015-06-14 | 2019-07-16 | Facense Ltd. | Blood pressure measuring smartglasses |
US20190313915A1 (en) * | 2015-06-14 | 2019-10-17 | Facense Ltd. | Posture-adjusted calculation of physiological signals |
WO2017129495A1 (en) * | 2016-01-28 | 2017-08-03 | Koninklijke Philips N.V. | Pulse rate measurement module and method |
CN108932452A (en) * | 2017-05-22 | 2018-12-04 | 中国科学院半导体研究所 | Arrhythmia cordis beat classification method based on multiple dimensioned convolutional neural networks |
US20200138306A1 (en) * | 2018-11-02 | 2020-05-07 | Samsung Electronics Co., Ltd. | Feature selection for cardiac arrhythmia classification and screening |
CN109635291A (en) * | 2018-12-04 | 2019-04-16 | 重庆理工大学 | A kind of recommended method of fusion score information and item contents based on coorinated training |
US20200222607A1 (en) * | 2019-01-16 | 2020-07-16 | Abiomed, Inc. | Left Ventricular Volume and Cardiac Output Estimation Using Machine Learning Model |
WO2020227514A1 (en) * | 2019-05-08 | 2020-11-12 | Lifelens Technologies, Llc | Monitoring and processing physiological signals to detect and predict dysfunction of an anatomical feature of an individual |
CN112071419A (en) * | 2019-06-11 | 2020-12-11 | 西门子医疗有限公司 | Hemodynamic analysis of blood vessels using recurrent neural networks |
CN111110208A (en) * | 2019-12-13 | 2020-05-08 | 南京理工大学 | LSTM-based oxygen reduction state prediction method for chronic obstructive pulmonary disease |
CN111493855A (en) * | 2020-04-21 | 2020-08-07 | 重庆理工大学 | System and method for non-invasive measurement of individualized cardiac output |
CN111528814A (en) * | 2020-04-29 | 2020-08-14 | 浙江工业大学 | Method for monitoring blood pressure through machine learning based on LSTM neural network |
CN112274127A (en) * | 2020-10-28 | 2021-01-29 | 河北工业大学 | Noninvasive continuous blood pressure detection method and device based on one-way pulse wave |
Non-Patent Citations (6)
Title |
---|
EVERSON, L; BISWAS, D; (...); VAN HELLEPUTTE, N: "BiometricNet: Deep Learning based Biometric Identification using Wrist-Worn PPG", 《IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS》 * |
THIELE R H, BARTELS K, GAN T J.: "Cardiac output monitoring: a contemporary assessment and review", 《CRITICAL CARE MEDICINE》 * |
XIAO, HG; BUTLIN, M; (...); AVOLIO, AP: "Estimation of Pulse Transit Time From Radial Pressure Waveform Alone by Artificial Neural Network", 《IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS》 * |
何为,肖汉光,刘兴华: "上肢动脉系统的三段式电网络建模与仿真", 《重庆大学学报》 * |
王丹妮: "基于深度学习的生物阻抗信号心功能评估方法研究", 《中国优秀硕士学位论文全文数据库》 * |
郑嘉强,程云章,边俊杰: "基于脉搏波特征参数的无创连续血压测量研究进展", 《中国医学物理学杂志》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113499048A (en) * | 2021-07-22 | 2021-10-15 | 重庆理工大学 | Central arterial pressure waveform reconstruction system and method based on CNN-BilSTM |
Also Published As
Publication number | Publication date |
---|---|
CN113057617B (en) | 2022-08-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102144916B (en) | Multi-channel pulse signal detecting method and device capable of automatically regulating pressure | |
CN101703396B (en) | Radial artery pulse wave based cardiovascular function parameter detection and analysis method and detection device | |
EP3358485A1 (en) | General noninvasive blood glucose prediction method based on timing analysis | |
CN103054562A (en) | Cardiovascular function detection method based on multi-channel pulse wave form analysis and device thereof | |
CN104068841B (en) | A kind of measuring method and device measuring Indices of Systolic Time parameter | |
CN104000573A (en) | Body surface two point pulse wave based central arterial pulse monitoring system and method | |
CN113143230B (en) | Peripheral arterial blood pressure waveform reconstruction system | |
CN112806977B (en) | Physiological parameter measuring method based on multi-scale fusion network | |
CN202960481U (en) | Traditional Chinese medicine pulse condition acquisition device | |
CN202920160U (en) | Traditional Chinese medicine pulse condition collection system | |
CN111839488B (en) | Non-invasive continuous blood pressure measuring device and method based on pulse wave | |
CN112971797A (en) | Continuous physiological signal quality evaluation method | |
CN103040524B (en) | Device and method for reducing interference of physiological activities to medical imaging or measuring results | |
CN113057617B (en) | Non-invasive monitoring system for cardiac output | |
CN109036552A (en) | Tcm diagnosis terminal and its storage medium | |
Yen et al. | Blood Pressure and Heart Rate Measurements Using Photoplethysmography with Modified LRCN. | |
US20150182140A1 (en) | Arterial pulse analysis method and system thereof | |
CN114145725B (en) | PPG sampling rate estimation method based on noninvasive continuous blood pressure measurement | |
CN115089145A (en) | Intelligent blood pressure prediction method based on multi-scale residual error network and PPG signal | |
CN109147945A (en) | Chinese Medicine Diagnoses System and bracelet | |
CN110811673A (en) | Heart sound analysis system based on probabilistic neural network model | |
CN107184194A (en) | Based on numerically controlled blood pressure self-operated measuring unit and method | |
CN113598734A (en) | Cuff-free blood pressure prediction method based on deep neural network model | |
CN208524857U (en) | Based on numerically controlled blood pressure self-operated measuring unit | |
Moukadem et al. | Automatic heart sound analysis module based on Stockwell transform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |